Biostatistics and Medical Informatics Department Seminar - DeMets Lecture with Janet Wittes

Dr. Wittes will discuss some of the common causes of non-proportionality in time to event clinical trials, focusing on three classes of reasons: 1 - experimental intervention leads to early harm or late benefit; 2. the study population at randomization is at non-constant risk and the time to event tends to be shorter in those of higher risk status; 3. a new standard of care introduced during the trial, or a pandemic that affects the two treatment groups differently.